Background Although serum ferritin measurement is the investigation of choice in iron deficiency, many laboratories continue to offer iron and TIBC/transferrin measurements. With increasing use of direct transferrin measurement, the possibility of reporting transferrin concentration alone without iron or transferrin saturation arises. This study compared the diagnostic utility of iron, transferrin and transferrin saturation measurements in the diagnosis of iron deficiency Methods Details of all laboratory requests for simultaneous serum iron, transferrin/TIBC and ferritin measurement were analysed to assess the diagnostic performance of iron, transferrin or TIBC, and transferrin saturation or TIBC saturation in identifying unequivocal iron deficiency (using 2 different serum ferritin cutoffs: < 12 and < 15 μg/l). The dataset was divided into 3 groups based on various reagent combinations. Results Across inpatient and outpatient populations and all 3 reagent combinations, transferrin or TIBC measurement outperformed iron measurement and saturation index. Mean areas under ROC curves across the study were: transferrin or TIBC: 0.94; Fe 0.77; saturation 0.87. There was no difference in diagnostic performance between transferrin and the 2 TIBC formulations. Conclusions Transferrin or TIBC measurement outperforms iron and saturation in predicting iron deficiency. This approach offers a cost-effective, evidence-based approach to the investigation of iron deficiency.